Abstract
Background:
Although empathic responding is considered a core competency in specialty palliative care (PC), patterns of empathic communication in PC encounters are not well understood.
Objectives:
In this secondary analysis, we delineate types and frequency of empathic communication and examine relationships between patient empathic opportunities and clinician responses.
Design:
We used the Empathic Communication Coding System to analyze empathic opportunities across three types: emotion (i.e., negative affective state), progress (i.e., stated recent positive life event or development), and challenge (i.e., stated problem or recent, negative life-changing event) and clinician responses.
Setting/Subjects:
Transcripts from a pilot randomized trial of communication coaching in specialty PC encounters (N = 71) audio-recorded by 22 PC clinicians at two sites in the United States: an academic health system and a community-based hospice and PC organization.
Results:
Empathic opportunities were frequent across encounters; clinicians often responded empathically to those opportunities (e.g., confirming or acknowledging patients' emotions or experiences). Even though challenge empathic opportunities occurred most frequently, clinicians responded empathically more often to progress opportunities (i.e., 93% of the time) than challenge opportunities (i.e., 75% of the time). One in 12 opportunities was impeded by the patient or a family member changing the topic before the clinician could respond.
Conclusions:
PC patients frequently express emotions, share progress, or divulge challenges as empathic opportunities. Clinicians often convey empathy in response and can differentiate their empathic responses based on the type of empathic opportunity. PC communication research and training should explore which empathic responses promote desired patient outcomes.
Introduction
Evidence-based, high-quality serious illness communication involves skilled navigation of discussions within specific encounter settings, 1 including responding to patients with empathy.2–4 Empathic communication is considered a core competency for fostering patient emotional stability, particularly in palliative care (PC) settings, where clinicians and patient discuss difficult topics.5–11 When empathic opportunities arise, clinicians can respond empathically to both acknowledge and process patients' disclosures and keep discussions relevant to clinician and patient encounter goals.12–14 Clinician empathic communication leads to reduced patient anxiety, enhanced care satisfaction, and patient emotional needs being met.9,10,15,16
Although clinician empathy in serious illness care has been identified as important, few researchers have studied how and to which opportunities PC clinicians respond with empathy. 12 For example, although researchers have assessed changes in empathy following PC communication training (e.g., dignity therapy, 17 motivational interviewing [MI], 18 and communication skill training19,20), few have examined patterns of empathic responding outside of training or coaching settings. Identification of processes of empathic communication between patients and clinicians in specialty PC provides guidance for defining the role of empathy, embedded in the context of multiple communication priorities in PC encounters, and identifying fitting goals for future PC communication interventions.
To systematically describe and analyze empathic communication in clinical encounters, researchers often use a process of (1) identifying empathic opportunities and then (2) coding the clinician's response to those opportunities. This study builds upon this process, using the Empathic Communication Coding System (ECCS) 21 to facilitate interaction analysis. The study has two aims. The first is to classify and describe empathic opportunities and empathic responses in specialty PC encounters. The second is to examine the relationship between empathic opportunity types and empathic response types in specialty PC encounters.
Methods
Setting and subjects
Our analyses were based on data from a pilot 2-arm randomized trial designed to teach MI skills to PC clinicians and were approved by the second author's Institutional Review Board (IRB # 202010037). 18 For the original study, clinicians were randomly assigned to an MI skill training intervention or waitlist control group. Group assignment was not examined in this study, as empathic opportunities and responses were examined across encounters, regardless of group. Data were collected from an academic community health system and in a community-based hospice and PC organization, both in the Mountain West, and both with established PC services. Full details about clinician and patient sample characteristics, informed consent, enrollment, feasibility, and primary outcomes of the MI intervention are described elsewhere. 18
Clinicians enrolled in the study were instructed to audio record four encounters with patients. The majority were initial appointments between patients and specialty PC clinicians or family meetings with multiple specialty PC clinicians present (e.g., physicians, nurses, social workers, and chaplains; range 1–6; mean [M] = 2.30; standard deviation [SD] = 1.49) and with patients' close others (i.e., range 0–9; M = 1.57; SD = 1.72) welcome to attend. Audio recordings were transcribed into written text by the original study team.
Design and methods
We coded transcripts of 71 encounters using the ECCS, 21 which was designed to identify empathic communication in physician–patient visits. The codebook outlines structured criteria for a two-step process of (1) identifying and coding empathic opportunities created by patients or close others during health care encounters and then (2) identifying empathic responses made by clinicians in direct response to empathic opportunities. Coders first identified distinct empathic opportunities, and then categorized clinicians' responses to each empathic opportunity into one of seven hierarchical categories of response.
Empathic opportunities
Empathic opportunities may begin when patients convey an emotion (i.e., fear, sadness, and anger). In this study, we considered both explicit statements of an emotion and indirect expressions of emotion (i.e., sighing and crying). Opportunities may also begin when patients verbalize progress (i.e., pertaining to recent, positive life developments or life-changing events) or challenge (i.e., describe the negative implications of an ongoing physical or psychological problem or recent, negative life-changing event). Empathic opportunities were identified and sorted into one of these three categories. In some cases, empathic opportunities could be described as both emotion and challenge opportunities. In these cases, opportunities were categorized as emotion opportunities, per original codebook guidelines. To create scores that control for encounter length, the frequency of each empathic opportunity type was divided by the total number of empathic opportunities in the encounter, resulting in proportion scores for each empathic opportunity type.
Empathic responses
Empathic responses refer to statements from a clinician that address an empathic opportunity and are characterized by seven distinct, hierarchical types further categorized into two groups. In this study, all responses from clinicians were considered in empathic response coding, regardless of the clinician's role. High-empathy responses explicitly acknowledge the empathic opportunity and include the following: shared experience (i.e., an explicit statement that the clinician shares the emotion or has had a similar experience), confirmation (i.e., validating or legitimizing the shared emotion or experience), acknowledgment with pursuit (i.e., conveying that the empathic opportunity was heard, with follow-up on the topic), and acknowledgment without pursuit (i.e., conveying that the empathic opportunity was heard, with no follow-up).
Low-empathy responses implicitly acknowledge or do not acknowledge the opportunity and include the following: implicit recognition (i.e., focusing on peripheral issues surrounding, but not central to, the empathic opportunity), perfunctory recognition (i.e., seemingly automatic, scripted, or minimal response to the empathic opportunity), and denial of perspective (i.e., clearly ignoring an empathic opportunity or making a disconfirming statement). An eighth type of response, following original codebook guidelines for “missing” responses, was labeled impeded response, and indicates that the clinician did not have an opportunity to respond empathically because the patient, close other, or another clinician changed the topic of conversation.
Empathic communication training and coding process
A team of two coders independently coded each transcript. Coder training involved careful review of a small subset of study transcripts and discussion of possible empathic opportunities and responses. Coders met weekly during training to practice coding and compare codes. Coders also developed refinements for the ECCS codebook to both improve clarity and revise criteria to fit the characteristics of PC encounters. Following the training period, coders independently coded one transcript for both empathic opportunities and responses, achieving good reliability (opportunities: Κ = 0.71 and responses: Κ = 0.73). Once reliability was established, both coders coded all study data and continued to meet weekly to resolve discrepancies in data and avoid coder drift.
Data analysis
SPSS version 27 was used to conduct statistical analyses. To address Aim 1, we calculated descriptive statistics for empathic opportunity and response study variables. To address Aim 2, we used chi-square to examine relationships between type of empathic opportunity (i.e., emotion, progress, and challenge) and empathic response (i.e., high-empathy response and low-empathy response). We then followed up with bivariate correlations between proportion of each type of empathic opportunity and frequency of low-empathy responses across encounters. Given the study's data were collected from occasions both before and after a communication skill-based training, we considered that relationships between empathic opportunities and empathic responses might differ across conditions and time. Follow-up and chi-squares were conducted encompassing the encounters that involved the waitlist control group, the intervention group at preintervention, and the intervention group at post-intervention, to explore possible differences across groups.
Results
Empathic opportunities
Coders detected 636 empathic opportunities across the 71 study encounters. All but one encounter contained at least one empathic opportunity. On average, encounters involved 9.08 empathic opportunities (SD = 5.51). Coders detected each type of empathic opportunity. Challenge opportunities were most frequent, occurring 319 times, or defining 50% of opportunities, and manifesting at least once in 94% of encounters. Emotion opportunities occurred 213 times (33% of opportunities), manifesting in 88% of encounters. Of emotion empathic opportunities, the most frequent emotion was fear (38% of emotion opportunities), followed by sadness (37%), anger (16%), and emotionally mixed (i.e., complex emotions that cannot be categorized as primarily fearful, angry, or sad; 9%). Progress empathic opportunities occurred least frequently, making up 105 opportunities (16%), and manifesting at least once in 60% of encounters. See Table 1 for exemplars of empathic opportunities initiated by patients.
Exemplars of Empathic Opportunities in Palliative Care Encounters
Empathic responses
All types of empathic responses delineated by the ECCS were found. Shared experience responses occurred 9 times (1%), confirmation 198 times (44%), acknowledgment with pursuit 42 times (9%), acknowledgment without pursuit 74 times (17%), implicit recognition 60 times (13%), perfunctory responses 16 times (4%) and denial of patient perspective 1 time (0.2%). Opportunities were impeded by the patient, close other, or another clinician 53 times (12% of responses). Thus, the majority of nonimpeded responses (80%) were classified as “high-empathy” responses, while only a small amount of nonimpeded responses (20%) were classified as “low-empathy” responses. See Table 2 for exemplars of empathic responses from PC clinicians.
Exemplars of Empathic Responses in Palliative Care Encounters
Relationship between empathic opportunities and responses
The chi-square analysis was significant, χ 2 = 29.09, p < 0.001, and Cramer's V = 0.21. Clinicians responded empathically most often to progress opportunities (i.e., high-empathy responses to 93% of progress opportunities), followed by emotion responses (i.e., high-empathy responses to 89.6% of emotion opportunities), and then challenge responses (i.e., high-empathy responses to 74.9% of challenge opportunities). Follow-up assessment of when low-empathy responses occur through examination of empathic opportunity proportion scores across transcripts revealed that clinicians gave fewer low-empathy responses in encounters with proportionally more progress opportunities (r = −0.25, p < 0.05), and clinicians gave more low-empathy responses in encounters with proportionally more challenge opportunities (r = −0.29, p < 0.05). There was no relationship between proportion of progress opportunities and low-empathy responses (r = 0.11, p > 0.05).
Although examining empathic communication within intervention groups and time points (i.e., preintervention, post-intervention, and waitlist control) was not a study aim, we considered the possibility that relationships between empathic opportunities and responses could vary across group conditions. As such, follow-up chi square analysis results, accounting for preintervention, post-intervention, and waitlist control study groups, mirror results reported in Aim 2. Relationships between empathic opportunity and low-empathy responses for each are significant: waitlist group χ 2 = 6.77, p < 0.05; intervention group at preintervention χ 2 = 15.63, p < 0.001; and intervention group at postintervention χ 2 = 12.59, p < 0.01. In all analyses, percent of high-empathy responding was highest for progress opportunities, then emotion, and lowest for challenge. Thus, results can be interpreted similarly regardless of intervention arm.
Discussion
This study describes the exchange of empathic opportunities and responses between patients and clinicians within the context of specialty PC. Our interaction analysis captured iterations of emotion, progress, and challenge empathic opportunities that are characteristic of PC encounters. The analysis also identified clinician empathic responses addressing patient empathic opportunities. Study results shed light on empathy as a central feature of specialty PC communication: empathic opportunities from patients are prevalent and predominantly met with high-empathy responses from clinicians. Because patient empathic opportunities were so frequent, we were able to examine variation in clinician empathic response, determining that high-empathy responses are somewhat tethered to type of empathic opportunity the clinician is addressing. Reasons for variation in empathic responses across types of opportunities, and implications of this variation for patient outcomes are important areas of consideration.
Variation in high-empathy responses across types of opportunities
Findings demonstrate that when patients expressed an emotion, PC clinicians often responded empathically. Clinicians' high-empathy responses were distributed unequally, however, across types of empathic opportunities. Progress opportunities received high-empathy responses most frequently. This finding is consistent with previous work on empathic clinician communication. 21 Research suggests that clinicians tend to communicate empathy more readily when patients display positive affect, 22 indicating that clinicians may naturally respond with empathy to progress opportunities. PC clinicians may also recognize that progress opportunities represent significant or poignant declarations, and that high-empathy responses can foster psychologically beneficial life reflection, 23 meaning making, 24 or existential maturation 25 in reference to recent perspectives on their illness.
In other words, when patients share a progress empathic opportunity, the clinician may feel confident that a high-empathy response will be both psychologically productive and relatively unlikely to disrupt encounter goals. PC clinicians may be particularly poised to foster productive psychological processes occurring during family meetings, further motivating their empathic responses to this type of opportunity.
In contrast, challenge opportunities received high-empathy responses less often, despite their frequency, and despite the field's understanding that patients who present multiple challenges may be in most need of empathic care from clinicians. 26 One possibility is that low-empathy responses to challenge opportunities may in some cases be employed to help discussions stay solution focused, avoiding unproductive catastrophizing 27 or bitterness revival 28 in reference to recent negative experiences. We speculate that PC clinicians (as well as, perhaps, other serious illness clinicians) may sometimes respond to extreme or ruminative statements in ways that indirectly acknowledge patient challenges, while attempting to keep discussions productive, although other reasons for low-empathy responses to challenge opportunities are possible. Garnering PC clinician and patient perspectives on responses to disclosures of challenge may help guide best practices for empathic communication in encounters where patients are highly affected by recent life challenges.
Specialty PC empathy: What is next for research and practice?
Research on empathic communication in clinician–patient encounters often culminates in discussion of need for interventions to increase low baseline rates of empathy in encounters. Even in PC contexts, empathic communication is typically discussed in terms of enhancing clinician-acquired skills, which convey attentive and empathic concern for the patient.11,29,30 Patient–clinician empathic communication, although, is ultimately dependent on patients communicating empathic opportunities, providing chances for clinicians to convey empathic concern. This study shows that, across clinicians who received MI skill coaching and those who had not yet received coaching, PC encounters involved considerably frequent empathic opportunities, even compared to other serious illness encounter settings (e.g., M = 1.21 empathic opportunities in lung cancer encounters 26 and M = 1.80 patient-prompted empathic responses in Dignity Therapy sessions with cancer patients 17 ).
Compared to other serious illness communication settings, the structure of specialty PC encounters, the interview style of specialty PC clinicians (i.e., asking open-ended questions about patients' identities, fears, and values), and the inclusion of multiple family members and multiple specialized clinicians in encounters likely contribute to the high number of empathic opportunities initiated during encounters. Future work should consider possible cyclical effects of empathic communication in this setting, with high-empathy discussions early-on in encounters potentially “setting the stage” for increased patient expression of empathic opportunities throughout the encounter. Future research should also tease apart the implications of multiple close others and multiple clinicians engaging in empathic communication in specialty PC encounter settings.
During encounters, clinicians often communicated highly empathic responses. This is distinct from other medical encounters,31,32 including serious illness encounters,26,33 where high-empathy responses are less frequent. Indeed, considering the high rate of empathic communication in specialty PC encounters, interventions that seek to increase rates broadly may not necessarily foster enhanced patient outcomes. 19 Instead, the future of empathic communication research in specialty PC settings may lie in more nuanced examination of, for example, whether notably highly empathic encounters directly influence desired patient outcomes (e.g., patient care satisfaction, adherence to medical recommendations, and navigation of decisions to purse disease-directed care) and which types of empathic responses most readily foster patient-centered outcomes.
These next research steps are presented considering the potential for competing goals of specialty PC encounters: offering emotional stability through empathic responding, while keeping discussions relevant to addressing patient medical needs. Although high-empathy responses may be objectively empathic, patients perceive high empathy not only in the things clinicians say (i.e., direct validation of patient experiences) but also in the things they do (i.e., provision of tailored medical recommendations) during encounters.34,35 Examining relationships between objective empathy, patient-perceived empathy in relationship to types of empathic responding, 34 and other patient outcomes in specialty PC will support efforts to identify best practices for empathic communication in this setting.
Considering the notable consistency of highly empathic communication in both PC settings examined in this study, another important next step in communication research is to recognize factors that foster frequent, high-empathy communication in specialty PC and identify ways to translate these factors to other clinical settings, where appropriate, 36 including in other serious illness care settings. 37 Currently, it is unclear which factors motivate highly empathic responding in specialty PC settings (e.g., PC clinicians' training, encounter setting and structure, and communication traits of PC clinicians).
Identifying which factors play key roles in high-empathy PC communication may be an important first step in directing the structure of communication training and assessment for other serious illness settings. We also consider that the nature of PC encounters, with focus on nondisease-directed treatments and quality of life, may in itself encourage patients to express empathic opportunities; in this case, translation of highly empathic communication to other settings may be less practical.
Limitations
This study is limited in its examination of empathic communication by 22 clinicians across 71 encounters, embedded in a larger pilot intervention project. Although study clinicians led multiple patient encounters, analyses could not account for differences in empathic communication across clinicians. Clinicians were asked to choose clinical encounters to record. Selection may have inflated or deflated estimates of empathic opportunities in specialty PC, although variation in the number of empathic opportunities across encounters (range: 0–29) suggests appropriate variability.
Demographic information was not recorded or linked to patients or clinicians in this study, so examining relationships between demographic characteristics (e.g., gender, race, and age) across both groups and empathic communication was not possible. Because patient characteristics are related to clinician empathic communication, 38 and concordance of patient and clinician characteristics can dictate clinician communication style and patient satisfaction,39,40 assessment of the relationship of these factors to empathic communication in this setting is warranted.
Conclusion
Authentic patient engagement is a cornerstone of high-quality, patient-centered care.41,42 Disclosure of empathic opportunities in care encounters defines one central aspect of patient encounter engagement. This study demonstrates that empathic opportunities frequently occur in specialty PC encounters, and thus the clinician's acumen to respond to the patient with validation is essential.13,43 As best practices for serious illness care continue to be charted, 37 adoption of patient- and clinician-endorsed practices for empathic care and communication in specialty PC, and translation of these practices to other serious illness care settings, is paramount.
Footnotes
Funding Information
Funding for original study data collection was provided by the National Palliative Care Research Center. Dr. Mroz is supported by NIA Institutional Training Grant: T32AG019134.
Author Disclosure Statement
No competing financial interests exist.
